Download - LSBU Research Open

The Effect of E-Cigarette Advertisements and Anti-Smoking Messages on Explicit and
Implicit Attitudes towards Tobacco and E-Cigarette Smoking in 18-65 Year Olds: A
Randomised Controlled Study Protocol
Paula Booth1, Ian P. Albery1, Daniel Frings1
Email: [email protected]
Tel: +44 (0)20 7815 5460
1
London South Bank University, Division of Psychology, School of Applied Sciences, 103
Borough Road, London SE1 0AA
Key words: E-cigarettes; Attitudes; Tobacco; Advertisements; Anti-Smoking Messages
Word Count – 4,443
1
Abstract
Introduction: Since the advent of e-cigarettes, e-cigarette advertising has escalated and
companies are able to utilise marketing strategies which are not permissible for tobacco
products. Research into the effect of e-cigarette advertising on attitudes towards tobacco and
e-cigarettes is in its infancy.
To date, no research has compared indirect (implicit) measures of attitude towards ecigarettes with direct (explicit) measures. Furthermore, little consideration has been given to
how viewing on-line advertisements may have an effect on attitudes towards e-cigarettes or
how positive attitudes to e-cigarettes may undermine anti-smoking public health messages.
The objectives of this study are to investigate (i) the relationship between explicit and
implicit attitudes towards tobacco and e-cigarettes, (ii) the effect of e-cigarette advertising on
these attitudes and (iii) the effect of these attitudes upon the efficacy of anti-smoking health
messages.
Methods and Analysis: In Experiment 1 an ANCOVA will be conducted to determine
whether viewing an e-cigarette advertisement, compared to a neutral image, has an effect on
implicit or explicit attitudes towards tobacco and e-cigarettes and if these attitudes differ
between smokers, vapers and non-smokers aged between 18 and 25 years. In Experiment 2,
moderation analysis will be conducted to assess whether attitudes towards e-cigarettes
moderate the psychological efficacy of anti-smoking health messages in participants aged 18
to 65 years. In each experiment, attitudes will be measured pre and post intervention and 1
week later (n=150), in participants that are smokers (n=50), vapers (n=50) or non-smokers
(n=50).
Ethics and Dissemination: Approval for this study has been given by the London South Bank
University’s’ (LSBU) Research Ethics Committee. The findings of these studies will be
submitted for publication and disseminated via conferences. The results will be integrated
into course provision for practitioners training at LSBU.
STRENGTHS AND LIMITATIONS

This study will determine whether positive attitudes towards e-cigarettes may
undermine anti-smoking campaign efficacy in smokers.

This study will be one of the first to determine whether direct measures of attitudes to
e-cigarettes relate to indirect measures in smokers, vapers and non-smokers.
2

This study is being carried out in a carefully controlled laboratory environment and as
such there is a lack of ecological validity.
INTRODUCTION
In the UK there are currently 2.8 million people using e-cigarettes[1] and it was estimated
that in 2014 projected sales in the USA would be worth over $2 billion.[2] E-cigarettes are
devices that deliver a vegetable glycerine and/or propylene glycerol vapour to a person when
they inhale or press a button on the device. The devices are battery operated and the vapour is
supplied in a cartridge or tank which generally, but not always, contains a varying amount of
nicotine.[3] The act of inhalation mimics smoking a cigarette and first generation e-cigarettes
looked very much like tobacco cigarettes. Over time the appearance of e-cigarettes have
evolved and become less like a traditional tobacco cigarette.[4] Unlike regular cigarettes, ecigarettes do not contain tobacco thus reducing smoking-related health risks.[5] However,
there are concerns that e-cigarettes may be harmful to health.[4] E-cigarettes have been
marketed as a smoking cessation device,[6] although evidence to support this notion is
inconsistent.[2] Additionally, there are concerns that non-smokers may be attracted to
smoking e-cigarettes and may act as a gateway to both nicotine addiction[7] and tobacco
use.[2]
Literature suggests that attitudes towards e-cigarettes are favourable.[8] Using e-cigarettes in
public places is reported as more acceptable than smoking tobacco cigarettes[9] and ecigarettes are perceived as safer,[3,8,10] healthier[8,10] and less addictive than tobacco
cigarettes.[10] Typically, e-cigarettes are considered a successful smoking cessation aid[10],
although young adults also consider e-cigarettes to be appealing to non-smokers.[11]
The studies above have examined explicit, rather than implicit, attitudes towards e-cigarettes.
Measures of explicit attitudes generally use direct self-report measures (e.g. surveys,
questionnaires or interviews) and may be influenced by social desirability biases questioning
the reliability of explicit attitudes as indicators of future behaviour.[12] For example,
participants may be reluctant to express positive attitudes towards tobacco cigarettes because
smoking tobacco cigarettes is a stigmatized behaviour and is not socially acceptable.[12]
However, health-related behaviours are also influenced by implicit attitudes.[13–15] Implicit
attitudes refer to favourable/unfavourable thoughts, feelings and behaviours towards an
attitude object which are measured indirectly and reflect immediate, impulsive reactions
rather than consciously controlled thought.[16-18] The role of implicit and explicit attitudes
3
in health-related behaviour is still debated as is whether the two mechanisms are mutually
exclusive of one another in predicting ongoing behaviour.[16–20]
In general, negative explicit attitudes are expressed towards tobacco smoking by both
smokers and non-smokers.[21] However, for implicit measures smokers have a more positive
reaction towards stimuli associated with smoking than non-smokers.[22,23] Other research
has found that smokers have a negative implicit attitude to smoking but that it is less negative
than for non-smokers.[21,24–26] This suggests that smokers are more likely to show
favourable attitudes towards smoking when reacting immediately and impulsively as opposed
to in a more controlled conscious manner.
Various tests have been developed to capture implicit attitudes (e.g. the evaluative priming
task, the semantic priming task, the Go/NoGo task and the single category Implicit
Association Test (SC-IAT)).[27] The SC-IAT is an established method of measuring implicit
associations between a single target (e.g. a cigarette) and two opposing attributes (bad and
good.[28] The implicit attitude is assessed by measuring the speed with which participants
categorise a target into one of the attribute categories. The test assumes that participants will
react more quickly if they automatically associate the target category with the attribute
category. For example, it would be assumed that a non-smoker would react more quickly if a
cigarette word was paired with a bad word, such as ‘horrible’, than if it were paired with a
good word, such as ‘love’.
This study will utilise SC-IATs with the personal attribute categories (SC-IAT-P) ‘I like’ and
‘I dislike’ rather than impersonal attributes, such as ‘good’ and ‘bad’, because evidence
suggests that participants may be influenced by societal norms when the categories are
impersonal.[23] For instance, Bardin et al. found that there was no significant difference
between smokers and non-smokers when labelling the attribute categories as ‘pleasant’ and
‘unpleasant; both smokers and non-smokers had negative implicit attitudes towards
cigarettes’. However, when labelling the attribute category as ‘I like’ and I dislike’ smokers
scores were neutral in comparison to the negative score for non-smokers and there was a
significant difference between them.[24] The design of the SC-IAT-Ps will be adapted from
that used in Bardin.[24]
Objective One: Relationships Between Attitudes
Since little research has investigated implicit attitudes towards e-cigarettes, the first objective
will examine the relationship (or dissociation) between e-cigarette-related implicit and
4
explicit attitudes. Furthermore, it has yet to be determined whether implicit or explicit
attitudes towards e-cigarettes may be associated with attitudes towards tobacco cigarettes.
Positive attitudes towards e-cigarettes may be related to either enhanced, depressed or neutral
attitudes towards tobacco cigarettes and these effects may be dependent on whether the
attitudes towards e-cigarettes are experienced in a consciously controlled way, impulsively or
both. Additionally, any association may be moderated by smoking group (smokers/nonsmokers/vapers).
Objective Two: Effects of E-Cigarette Advertisements
The second objective tests whether viewing on-line e-cigarette advertising affects implicit
and explicit attitudes towards e-cigarettes and tobacco cigarettes in young people.
Consistently, research has shown that viewing e-cigarette advertisements is likely to increase
positive attitudes and intentions to try e-cigarettes. In a group of adolescents (13-17 years),
who had never tried e-cigarettes, but who were either tobacco smokers or non-smokers, those
who answered a survey after viewing an e-cigarette advert had a more positive attitude
towards e-cigarettes, scored more highly on the potential benefits of e-cigarettes and had a
significantly higher intention of trying e-cigarettes than those who answered the survey
before being exposed to an e-cigarette advert. Additionally, current smoking of tobacco
cigarettes increased the likelihood of using e-cigarettes in the future.[29] Other studies
comprising adult smokers[30] and non-smokers[31] report increased intention or more
openness for trying e-cigarettes after viewing an advertisement. Conversely, Nagelhout et al.
study found there was no correlation between noticing e-cigarette advertisements and ecigarette use.[32] However, in a study in which participants were asked to indicate if they
wished to try an e-cigarette or snus (smokeless tobacco product) after viewing magazine
adverts for either e-cigarettes or snus, only 16% of participants indicated that they intended to
try the product.[33] When they were offered a voucher for a free product at the end of the
study 45% asked for a voucher. This discrepancy may arise because participants did not want
to self-report a wish to try the product and that the voucher request may be a more reliable
measure of intention to try. Alternatively, participants may have had the intention of getting
the free product to give to a friend/family member. It is predicted that viewing an e-cigarette
advertisement will increase positive attitudes towards e-cigarettes in smokers, vapers and
non-smokers both at an implicit and explicit level.
5
Findings for the effects of e-cigarette advertisements on attitudes towards tobacco smoking
are less consistent. In one study, results showed that viewing an e-cigarette advertisement
may encourage tobacco smokers to smoke more. Significantly more daily smokers had a
tobacco cigarette whilst viewing an e-cigarette advertisement if it showed e-cigarette
imagery, than if the auditory track of the advertisement remained but the imagery was
replaced with words rather than pictures.[34] Conversely, Nagelhout et al. showed that
noticing e-cigarette advertisements, in tobacco cigarette smokers, was associated with greater
disapproval of tobacco smoking. However, there was no association between e-cigarette
advertisements and tobacco cigarette quit attempts/successes.[32] The current study will
examine the effect of viewing an e-cigarette advertisement on attitudes towards tobacco
smoking in young people. Younger people are the focus of this experiment as studies suggest
that young people are deliberately targeted by e-cigarette advertisements and that this cohort
is not generally using e-cigarettes as a smoking cessation tool.[35] It will determine whether
viewing an e-cigarette advertisement increases positive attitudes towards tobacco smoking in
smokers, vapers and non-smokers and whether these effects are found for explicit and
implicit beliefs.
Objective Three: The Effect of Attitudes Towards E-cigarettes on the Efficacy of AntiSmoking Messages
Literature suggests that mass media campaigns are effective in encouraging quit attempts and
reducing tobacco smoking in adults[36] but that the effect of anti-smoking campaigns on
adolescents and young people is less robust.[37] There is no published literature on whether
attitudes towards e-cigarettes may moderate the efficacy of anti-smoking campaigns. There is
a concern that favourable attitudes towards e-cigarettes may ‘spill over’ into attitudes towards
tobacco cigarettes and undermine the efficacy of anti-smoking campaigns. Alternatively,
having a positive attitude towards e-cigarettes may enhance the effectiveness of anti-smoking
campaigns and encourage people to use e-cigarettes as a smoking cessation tool. The third
objective will determine whether attitudes towards e-cigarettes moderate the effectiveness of
anti-smoking messages in smokers. The age group for this experiment will be broadened to
include 18 to 65 year olds as quit smoking campaigns are more effective in adults as opposed
to young people.[36,37]
This programme of research will investigate these three objectives across two experimental
studies by examining (i) the relationship between explicit and implicit attitudes towards e-
6
cigarettes and tobacco cigarettes, (ii) the effect of e-cigarette advertising on these attitudes
and (iii) the effect of attitudes towards e-cigarettes on the psychological efficacy of antihealth messages.
METHODS/DESIGN
Design: This study has an experimental design with measurements being taken from each
participant at baseline, immediately post-intervention and one week later (see Figure 1). Two
experiments are being undertaken: Experiment 1 uses an image of e-cigarette advertisement
as the intervention and a neutral image as the control; Experiment 2 uses an image of antismoking advertisement as the intervention and a neutral image as the control. Each group of
participants (smokers, non-smokers, e-cigarette users) is randomly allocated to either the
control or intervention group in a 1:1 ratio by an Excel-based covariate adaptive
randomization program.[38]
The dependent variables are the scores derived for explicit and implicit attitudes. Explicit
attitudes are being measured through a self-report questionnaire and have been previously
used to measure attitudes towards e-cigarettes.[39] Implicit attitudes are being measured
using two Single Category Personalised Implicit Association Tests (SC-IAT-P).
Participants
In Experiment 1 adults aged 18 to 25 years are eligible. In Experiment 2, the age range has
been broadened to include participants aged between 18 and 65 years. Participants must
either be current tobacco cigarette smokers (smoked at least 10 cigarettes in the last 30 days),
current e-cigarette users (vaped at least ten times in the last 30 days) or non-smokers (not
smoked or vaped in the last 30 days). Participants are not eligible to take part if they are dual
users (i.e. smoke both tobacco and e-cigarettes), pregnant (or suspect pregnancy), breastfeeding or have a serious medical condition. When potential participants enquire about taking
part in the study they are sent information listing these criteria. Participants are asked about
their age and smoking status when they sign up for the study and eligibility is re-checked at
the first testing session. Advertising for students within a university is the primary
recruitment method but advertising locally for participants is also necessary to ensure the
required sample size. Participation is voluntary and to encourage study retention all
volunteers are offered £30 on completion of all testing sessions. Alternatively, psychology
undergraduate students are offered twelve participation credits towards a required total as part
of their degree programme.
7
Procedure
The study is currently taking place in a London university single occupancy experimental
laboratory. One employed researcher and/or protocol trained intern is carrying out testing.
Data collection began in March 2016 and is scheduled to finish April 2017. Participants are
tested individually. For Experiment 1 participants must attend three testing sessions - a preintervention test session on day one, the intervention and post-intervention testing on day
three and a further post-intervention session on day ten. Data collection for Experiment 1
began first and feedback from participants suggested that recruitment might be less
challenging if participants had to attend on fewer occasions. Thus, in Experiment 2,
participants complete both pre- and post-intervention testing on day one and a further postintervention session on day seven.
Upon arrival at the lab participants are seated in front of a Viglen desktop computer (Viglen
Genie i5 650 Processor 3.2GHz, Viglen Ltd., UK) with a HX193DPB 19 inch LED Monitor.
Screening for eligibility for the study carried out by email before the testing day is then
confirmed. Participants are given a copy of the study information, which has already been
sent to them by email and written consent collected. First, participants are presented with the
SC-IAT-P task on the computer screen. Verbal instructions are given to the participant at the
beginning of the task and understanding checked such that the experimenter and participant
are content that they are able to respond correctly. This task takes 15-20 minutes. Secondly,
the explicit attitude task is presented on the screen. These two tasks are given pre- and postintervention and one week post-intervention. In the pre-intervention session only a smoking
history and beliefs questionnaire is given to the participant to complete after the explicit
attitude test. After completion of the final session participants are given a debriefing and
£30/study credits.
Intervention
Two pilot studies have been carried out to identify an image of both (i) an e-cigarette
advertisement and (ii) an anti-smoking message for use in Experiments 1 and 2.
E-Cigarette Advertisements
In pilot Study 1, 200 different on-line e-cigarette advertisement images displayed since 2013
were found. Ten themes were identified by the protocol author depicting e-cigarettes as: a
smoking cessation tool; a healthy alternative; satisfying as smoking; being able to be used in
8
places where the smoke free legislation applies; being cool and having cosmetic appeal;
being cleaner and more fragrant; being sporty; being cheaper than tobacco cigarettes; being
endorsed by celebrities and improving social relationships. Each theme was given a code and
every advertisement coded by the author. Coding was undertaken independently by four
research interns. Advertisements not coded consistently by at least three of the five coders
were discarded. From the reduced set of thirty-two advertisements the research team chose
fifteen advertisements based on which they found the most engaging. These images were
placed in a survey developed on Qualtrics to determine which advertisement was the most
persuasive and effective.
A link to the survey was placed on CrowdFlower (an on-line crowd-sourcing site which pays
contributors to complete on-line research). This survey required participants to rate how
persuasive, effective, happy, sad, scary, exciting and engaging they found the advertisements
on a scale between 0 (not at all) and 100 (very). Additionally, control questions were built in
which required participants to choose a specific rating to indicate that they were human and
were paying proper attention in the survey (e.g. “are you a proper person? Please rate this
answer as strongly agree.”) A total of 959 participants completed the survey. Data from
participants were excluded if participants were less than 18 years old or if they did not rate
the control questions. Responses from 776 participants were analysed to identify which
image had the highest mean score for each adjective. One advertisement was chosen for use
in Experiment 1 which had the highest score in terms of persuasiveness and effectiveness,
and which was not affected by gender and did not portray cigarette smoking negatively. A
control image was created by the protocol author that was very similar to the e-cigarette
advertisement but the image of the e-cigarette was replaced with an electronic toothbrush and
the wording altered to advertise a toothbrush rather than an electronic cigarette. The image
was evaluated and agreed by the research team.
Anti-smoking Health Messages
In pilot Study 2, 97 online anti-smoking messages were identified. Two interns identified 15
different themes: toxic gore; guns/suicide; ingredients; teen empowerment; ultimatum; lung
cancer; homicide; premature ageing; amputation; teeth; recovery; children; death; expense
and addiction. Each anti-smoking image was categorised as a theme and after discussion
between the interns one anti-smoking image from each theme was selected on the basis of
which image they considered to be the most engaging. An identical survey to that developed
9
in pilot Study 1 was used but the e-cigarette advertisement images were replaced with the
anti-smoking message images.
Again, a link to the survey on Qualtrics was placed on CrowdFlower. From the original 990
participants who answered the survey, 17 were excluded for being under 18 years old, 9
because the image number was not recorded in Qualtrics and 146 because they did not rate
the control questions correctly. The anti-smoking image with the highest scores for
persuasiveness and effectiveness was chosen as the image to be used in Experiment 2. The
anti-smoking message chosen showed that physical disabilities may be caused by smoking.
Three different control images were designed by the protocol author and evaluated by the
research team. The chosen image showed the same image as the test intervention but the
wording was changed to state that travelling on public transport may be difficult for people
with disabilities.
Measures
Experimental controls
The smoking history and beliefs survey was developed from previous literature. Current
usage and dependence on tobacco cigarettes is determined by use of the Fagerström Test for
Nicotine Dependence (FTND) and adapted for smokers of e-cigarettes and ex-smokers.[3]
The strength of beliefs about e-cigarettes is measured using a five point Likert Scale (strongly
agree to strongly disagree) and has been utilised previously.[37] Participants are asked to rate
the following statements; “using e-cigarettes can help people quit smoking”, “using ecigarettes is less harmful to the health of the user than smoking cigarettes”, and “e-cigarettes
are less addictive than cigarettes.” A scale to represent participant’s motivation to quit was
adapted from previous literature.[40] Additionally, the number of previous quit attempts will
be recorded. Awareness of e-cigarette advertising and anti-smoking campaigns will be
measured by adapting questions originally used in a study to measure receptivity of
participants to smoking advertisements.[41] Participants will be asked if they recall seeing an
e-cigarette advertisement recently on TV, in a magazine/newspaper or on a social media site
(e.g. Facebook). Results will be used as covariates to control for differences in smoking
history, beliefs and awareness of e-cigarette campaigns.
Explicit Attitudes to Tobacco and E-cigarettes.
10
The measure is delivered via Qualtrics, an online survey tool, and completed in the lab during
the testing session. Using seven point semantic differential scales from previous work[19,25],
participants are asked to indicate how bad/good; harmful/harmless; foolish/wise;
unpleasant/pleasant; boring/exciting; not enjoyable/enjoyable; sexy/not sexy;
healthy/unhealthy; glamorous/ugly and calming/stressful’ they think e-cigarettes/tobacco
cigarettes are. The order of the required ratings will be counter-balanced.
Implicit Attitudes to Tobacco and E-cigarettes.
Implicit attitude is being measured using two personalised SC-IAT-P tests (for e-cigarettes
and tobacco cigarettes). The labels ‘I like’ and ‘I dislike’ are presented on opposite sides at
the top right-hand and left-hand side of the screen. The label ‘tobacco cigarettes’ or ‘ecigarettes’ is presented directly under either the ‘I like’ or ‘I dislike’ label. Stimuli words,
attribute category words associated with ‘I like’, ‘I dislike’ or words associated with either
tobacco or e-cigarettes appear in the middle of the screen. The participant categorises the
stimuli word into one of the labels by pressing a key which represents either the left-hand
side (‘E’ key) or the right-hand side (‘O’ key). Stimuli appear on the screen until participants
respond or for a maximum of 1500 milliseconds (ms).[24] After this time a message
‘FASTER’ is presented for 500 ms.
Seven words associated with tobacco cigarettes were drawn from previous work [21,42]
tobacco, smoking, cigarettes, smokers, ashtray, lighter and matches. A pilot study of students
identified words that are were most strongly associated with e-cigarettes - vaporizer, vaping,
vaper, e-liquid and vape. Attribute category words were drawn from previous work [26]. For
the ‘I dislike’ category: torture, ugly, war, injury, funerals, pus, pain, infection, vomit and
sad; the ‘I like’ category: sweetness, holidays, happiness, flower, happy, victory, gifts,
festive, giving and beauty. Stimuli and labels are presented in white upper-case letters (5 mm
high, 4 mm wide) on a black background.[28]
Prior to each block an information screen guides participants which key will be associated
with which labels (see Figure 2). In each SC-IAT-P (separately for tobacco and e-cigarettes)
there are four stages of 4 blocks (24 word trials per block); there are 24 practice trials and 72
test trials in each stage. Table 1 shows the order of label presentation. In stages 1 and 4 words
are presented in a 7:7:10 ratio - tobacco/e-cigarette words plus ‘I like’ words vs ‘I dislike’
words and in stages 2 and 3 a ratio of 7:10:7 is used. Millisecond reaction times and errors
are recorded.
11
Table 1
Presentation of labels in the SC-IAT-P
Stage
Trials
Function
‘E’ key response
‘O’ key response
1
24
Practice
I like words +
I dislike
72
Test
tobacco cigarettes
24
Practice
I like
72
Test
24
Practice
I dislike words +
72
Test
tobacco cigarettes
24
Practice
I dislike
72
Test
2
3
4
I dislike words +
tobacco cigarettes
I like
I like words +
tobacco cigarettes
The algorithm used by Bardin et al.[24] and Karpinski[43] based on the D-score from
Greenwald et al. will be used when analysing the data.[44] Responses from practice blocks,
non-responses, those less than 350 msec and from participants with 20% or more errors will
be removed. Error responses will be replaced with the sum of the mean block time plus a 400
millisecond penalty. To determine whether participants more strongly (quickly) associate
tobacco/e-cigarettes with like or dislike, the average response time for ‘I like’ will be
subtracted from ‘I dislike’. These scores will be divided by the pooled standard deviation of
all test trial times to give a standardised D score. Higher scores will indicate a more positive
attitude towards smoking/vaping. Internal consistency of the P-IAT-SCs will be assessed
using the procedure described by Frings et al.[45]
STATISTICAL ANALYSIS
Sample Size Analysis
Based on G-Power calculation in each experiment 150 participants will be required (50 ecigarette users, 50 tobacco cigarette smokers and 50 non-smokers). With three time point
measures and a sample of 150 spread across six groups (smoker type x intervention
conditions) sufficient statistical power (95%) will be present to detect a small effect size (f
=.17) in a repeated measures interaction. An ANCOVA between-groups with six groups, no
time within-subjects factor and 3 covariates (chosen from experimental controls measures),
will be sufficiently powered (95%) to detect a medium size effect (f =.32).
12
Statistical Analysis Plan
All data for participants who complete part one, but not part two and three, of the study will
be excluded. As recommended by SPIRIT a full description of missing data in each condition
and smoking group will be provided.[46] If the attrition rate, in either condition or overall,
exceeds 20% for Time 3 then all Time 3 data will be removed from analysis.[47] If attrition is
less than 20% missing data will be replaced with the ‘worst ranking score’ as the most
conservative method.[48]
Data Preparation
Data will be checked to confirm that they meet the assumptions for Pearson’s correlation
analysis (alternative analyses will be considered if violations occur). Before analysing data
using the Hayes PROCESS model assumptions for multiple regression will be checked for
multivariate normality and linearity (transformation considered if assumptions not met) and
Homoscedasticity. For multicollinearity between independent variables if rs > 0.7 one of the
offending variables will be removed.
Before ANOVA data for each outcome at Time 1, 2 and 3 will be assessed in terms of
linearity and transformation considered if violated. If normality is violated a non-parametric
test will be used. Levene’s test and Hartley’s F max test will check homogeneity of variance.
Greenhouse-Geisser statistics will be used if data violates assumptions of sphericity. Before
conducting ANCOVA covariates and explicit belief variables will be checked for
multicollinearity. If any rs >.07 one of the offending variables will be excluded. Analyses
conducted with covariate and non-covariate models will be reported.
Primary Outcome Measures
Objective 1:Relationships between attitudes
Correlational analyses will determine whether there is a relationship between explicit and
implicit attitudes towards e-cigarettes at baseline. Additionally, the association between
attitudes towards e-cigarettes and tobacco cigarettes will be investigated and moderation
analysis will test the observed relationships within each smoking group.
Objective 2: Effect of advertising on attitudes
To determine whether viewing an advertisement has an effect on attitudes, mixed model
ANOVAs will be conducted with TIME (pre-intervention/post-intervention/1 week) as a
13
within-subjects factor and INTERVENTION (advertisement/control) and SMOKING
GROUP (smokers/vapers/non-smokers) as between-subjects factor. In each experiment, one
ANOVA will be conducted with implicit outcome measures as the dependent variable, and a
second with the explicit test outcomes. This analysis will be repeated for attitudes towards
tobacco and e-cigarettes.
Objective 3: The effect of attitudes on anti-smoking health message psychological efficacy
A moderation analysis using Hayes PROCESS model will determine whether attitudes
towards e-cigarettes at baseline moderate the psychological efficacy of anti-smoking health
messages on implicit/explicit attitudes towards tobacco cigarettes. Baseline tobacco scores
will be co-varied and the moderating effect of smoking groups will be investigated: the
smoking group data will be recoded so that non-smoker’s data will be compared to first
vapers and secondly smokers.
Subsidiary Analysis
Baseline current usage/dependence upon tobacco/e-cigarettes, beliefs about e-cigarettes,
motivation to quit, number of previous quit attempts, awareness of e-cigarette advertising and
anti-smoking campaigns will be added as covariates. Additionally, a subsidiary analysis will
be carried out to determine whether there is an effect of age on implicit and explicit attitudes
towards tobacco and e-cigarettes both before and after viewing either an e-cigarette
advertisement or an anti-smoking message.
ETHICS AND DESSIMINATION
Informed consent will be collected prior to testing. It is intended that the dataset will be made
available publicly via UK data repository services. Findings will be disseminated via peerreviewed publications, scientific conferences and training events (e.g. CPD activities). A nontechnical summary will be offered to Cancer Research UK and policy/pressure organisations
(e.g. Action on Smoking and Health, Action on Addiction). Ethical approval has been granted
from the London South Bank University Research Ethics Committee (UREC number 1602).
CONCLUSION
This study will assess the relationship between direct and indirect measures of attitudes
towards e-cigarettes and tobacco cigarettes. No known studies have previously investigated
differences in implicit and explicit attitudes towards e-cigarettes. This study will show if
14
viewing the image of an e-cigarette advertisement changes attitudes, and whether attitudes
and attitude changes differ between smokers, non-smokers and e-cigarette users.
Furthermore, this is the first study to consider whether positive attitudes towards e-cigarettes
may potentially undermine public health campaigns encouraging smokers to quit. Findings
will be important in informing policy around existing concerns about the relationship
between e-cigarette attitudes and tobacco cigarettes, advertising and anti-smoking health
campaigns. It will also inform practitioners (e.g. GPs, cessation counsellors, etc.) by
providing insight into the role of e-cigarettes in smoking cessation.
REFERENCES
1
ASH Action on Smoking and Health. Use of electronic cigarettes (vapourisers) among
adults in Great Britain. ASH Fact Sheet 2016. 2016.
2
Schraufnagel DE, Blasi F, Drummond MB, et al. Electronic cigarettes. A position
statement of the forum of international respiratory societies. Am J Respir Crit Care
Med 2014;190:611–8. doi:10.1164/rccm.201407-1198PP
3
Dawkins L, Turner J, Roberts A, et al. ‘Vaping’ profiles and preferences: an online
survey of electronic cigarette users. Addiction 2013;108:1115–25.
doi:10.1111/add.12150
4
Rom O, Pecorelli A, Valacchi G, et al. Are E-cigarettes a safe and good alternative to
cigarette smoking? Ann N Y Acad Sci 2015;1340:65–74. doi:10.1111/nyas.12609
5
McNeill A et al. E-cigarettes: An Evidence Update. A Report Commissioned by Public
Health England. 2015.
6
Zhu S-H, Sun JY, Bonnevie E, et al. Four hundred and sixty brands of e-cigarettes and
counting: implications for product regulation. Tob Control 2014;23 Suppl 3:iii3-9.
doi:10.1136/tobaccocontrol-2014-051670
7
Durbin RJ, Waxman HA, Harkin T, Rockefeller JD, Blumenthal R ME. Gateway to
addiction? A survey of popular electronic cigarette manufacturers and targeted
marketing to youth. InUS Congr;711.
8
Pepper JK, Brewer NT. Electronic nicotine delivery system (electronic cigarette)
awareness, use, reactions and beliefs: a systematic review. Tob Control 2014;23:375–
84. doi:10.1136/tobaccocontrol-2013-051122
9
Trumbo CW, Harper R. Orientation of US Young Adults toward E-cigarettes and their
Use in Public. Heal Behav Policy Rev 2015;2:163–70.
10
Choi K, Forster J. Characteristics associated with awareness, perceptions, and use of
electronic nicotine delivery systems among young US Midwestern adults. Am J Public
Health 2013;103:556–61. doi:10.2105/AJPH.2012.300947
11
Choi K, Fabian L, Mottey N, et al. Young adults’ favorable perceptions of snus,
dissolvable tobacco products, and electronic cigarettes: findings from a focus group
study. Am J Public Health 2012;102:2088–93. doi:10.2105/AJPH.2011.300525
15
12
Sherman SJ, Chassin L, Presson C, et al. The Intergenerational Transmission of
Implicit and Explicit Attitudes Toward Smoking. J Exp Soc Psychol 2009;45:313.
doi:10.1016/j.jesp.2008.09.012
13
Sheeran P, Gollwitzer PM, Bargh JA. Nonconscious processes and health. Heal
Psychol 2013;32:460–73. doi:10.1037/a0029203
14
Gawronski B, Bodenhausen G V. Associative and propositional processes in
evaluation: An integrative review of implicit and explicit attitude change. Psychol Bull
2006;132:692–731. doi:10.1037/0033-2909.132.5.692
15
Wiers, R. W., Houben, K. M. P. I., Roefs, A., Jong, P., Hofmann, W., & Stacy AW.
‘Implicit cognition in health psychology: why common sense goes out of the window.’
In: Gawronsk B& PBK, ed. Handbook of implicit social cognition. New York, NY,
US: : The Guilford Press 2010. 463–88.
16
Rooke SE, Hine DW, Thorsteinsson EB. Implicit cognition and substance use: a metaanalysis. Addict Behav 2008;33:1314–1328 15p.
17
Roefs A, Huijding J, Smulders FTY, et al. Implicit measures of association in
psychopathology research.
18
Frings D, Albery IP. The Social Identity Model of Cessation Maintenance: formulation
and initial evidence. Addict Behav 2015;44:35–42. doi:10.1016/j.addbeh.2014.10.023
19
Perugini M. Predictive models of implicit and explicit attitudes. Br J Soc Psychol
2005;44:29–45. doi:10.1348/014466604X23491
20
Reich RR, Goldman MS, Noll JA. Using the False Memory Paradigm to Test Two
Key Elements of Alcohol Expectancy Theory.
21
Huijding J, de Jong PJ, Wiers RW, et al. Implicit and explicit attitudes toward smoking
in a smoking and a nonsmoking setting. Addict Behav 2005;30:949–61.
doi:10.1016/j.addbeh.2004.09.014
22
Huijding J, de Jong PJ. Automatic associations with the sensory aspects of smoking:
positive in habitual smokers but negative in non-smokers. Addict Behav 2006;31:182–
6. doi:10.1016/j.addbeh.2005.04.014
23
De Houwer J, Custers R, De Clercq A. Do smokers have a negative implicit attitude
toward smoking? Cogn Emot 2006;20:1274–84. doi:10.1080/02699930500484506
24
Bardin B, Perrissol S, Py J, et al. Personalized SC-IAT: a possible way of reducing the
influence of societal views on assessments of implicit attitude toward smoking.
Psychol Rep 2014;115:13–25. doi:10.2466/18.07.PR0.115c10z8
25
Swanson JE, Swanson E, Greenwald AG. Using the Implicit Association Test to
investigate attitude-behaviour consistency for stigmatised behaviour. Cogn Emot
2001;15:207–30. doi:10.1080/02699930125706
26
Robinson M, Meier B. Smoking and the Implicit Association Test: When the contrast
category determines the theoretical conclusions. Basic Appl … 2005.
27
Nosek BA, Hawkins CB, Frazier RS. Implicit social cognition: from measures to
mechanisms. Trends Cogn Sci 2011;15:152–9. doi:10.1016/j.tics.2011.01.005
28
Karpinski A, Steinman RB. The Single Category Implicit Association Test as a
16
measure of implicit social cognition.
29
Farrelly MC, Duke JC, Crankshaw EC, et al. A Randomized Trial of the Effect of Ecigarette TV Advertisements on Intentions to Use E-cigarettes. Am J Prev Med
2015;49:686–93. doi:10.1016/j.amepre.2015.05.010
30
Kim AE, Lee YO, Shafer P, et al. Adult smokers’ receptivity to a television advert for
electronic nicotine delivery systems. Tob Control 2015;24:132–5.
doi:10.1136/tobaccocontrol-2013-051130
31
Popova L, Ling PM. Nonsmokers’ responses to new warning labels on smokeless
tobacco and electronic cigarettes: an experimental study. BMC Public Health
2014;14:997--997 1p. doi:10.1186/1471-2458-14-997
32
Nagelhout GE, Heijndijk SM, Cummings KM, et al. Noticing e-cigarette
advertisements and associations with use of e-cigarettes, disapproval of smoking, and
quitting smoking. Findings from the International Tobacco Control (ITC) Netherlands
Survey. Int J Drug Policy Published Online First: December 2015.
doi:10.1016/j.drugpo.2015.12.015
33
Smith DM, Bansal-Travers M, O’Connor RJ, et al. Associations between perceptions
of e-cigarette advertising and interest in product trial amongst US adult smokers and
non-smokers: results from an internet-based pilot survey. Tob Induc Dis 2015;13:14.
doi:10.1186/s12971-015-0039-6
34
Maloney EK, Cappella JN. Does Vaping in E-Cigarette Advertisements Affect
Tobacco Smoking Urge, Intentions, and Perceptions in Daily, Intermittent, and Former
Smokers? Health Commun 2015;:1–10.
35
De Andrade M, Hastings G, Angus K, et al. the marketing of electronic cigarettes in
the uk. Published Online First:
2013.https://dspace.stir.ac.uk/bitstream/1893/17889/1/deAndradeetale-cigsreport.pdf
(accessed 26 Mar2017).
36
Durkin S, Brennan E, Wakefield M. Mass media campaigns to promote smoking
cessation among adults: an integrative review. Tob Control 2012;21:127–38.
doi:10.1136/tobaccocontrol-2011-050345
37
Choi K, Forster JL. Beliefs and experimentation with electronic cigarettes: A
prospective analysis among young adults. Am J Prev Med 2014;46:175–8.
doi:10.1016/j.amepre.2013.10.007
38
Kang, M., & Park JH. Covariate Adaptive Randomization Program. 2007.
39
Smith DM, Bansal-Travers M, O’Connor RJ, et al. Associations between perceptions
of e-cigarette advertising and interest in product trial amongst US adult smokers and
non-smokers: results from an internet-based pilot survey. Tob Induc Dis 2015;13:14.
doi:10.1186/s12971-015-0039-6
40
Zhou X, Nonnemaker J, Sherrill B, et al. Attempts to quit smoking and relapse: factors
associated with success or failure from the ATTEMPT cohort study. Addict Behav
2009;34:365–73. doi:10.1016/j.addbeh.2008.11.013
41
Gilpin EA, White MM, Messer K, et al. Receptivity to tobacco advertising and
promotions among young adolescents as a predictor of established smoking in young
adulthood. Am J Public Health 2007;97:1489–95. doi:10.2105/AJPH.2005.070359
17
42
Munafò M, Mogg K, Roberts S, et al. Selective Processing of Smoking-Related Cues
in Current Smokers, Ex-Smokers and Never-Smokers on the Modified Stroop Task. J
Psychopharmacol 2003;17:310–6. doi:10.1177/02698811030173013
43
Karpinski A, Steinman R. The single category implicit association test as a measure of
implicit social cognition. J Personal Soc … 2006.
44
Greenwald, A. G., Nosek, B. A., & Banaji MR. "Understanding and using the implicit
association test: I. An improved scoring algorithm. J Pers Soc Psychol 2003;85:197.
45
Frings D, Melichar L, Albery IP. Implicit and explicit drinker identities interactively
predict in-the-moment alcohol placebo consumption. Addict Behav Reports
2016;3:86–91. doi:10.1016/j.abrep.2016.04.002
46
Chan A-W, Tetzlaff JM, Gøtzsche PC, et al. SPIRIT 2013 explanation and elaboration:
guidance for protocols of clinical trials. BMJ 2013;346:e7586. doi:10.1136/bmj.e7586
47
Whitlock EP. Behavioral Counseling Interventions in Primary Care To Reduce
Risky/Harmful Alcohol Use by Adults: A Summary of the Evidence for the U.S.
Preventive Services Task Force. Ann Intern Med 2004;140:557. doi:10.7326/00034819-140-7-200404060-00017
48
Berger VW. Conservative handling of missing data. Contemp Clin Trials 2012;33:460.
doi:10.1016/j.cct.2012.02.008
Authors' contributions: Dr Frings and Prof. Albery conceptualized and designed the study and
critically reviewed and revised the manuscript; Dr Booth drafted the initial manuscript and
will administer the testing, data collection and statistical analysis. All authors approved the
final manuscript as submitted.
Funding statement: 'This work was supported by Cancer Research UK grant number
C54622/A20485’.
Competing Interest Statement: Daniel Frings and Ian Albery are both investigators on a
randomised controlled trial funded by Allen Carrs Easyway Ltd. This trial is comparing the
Allen Carr Easway stop-smoking method to local NHS 1-1 stop smoking counselling service.
The team are contractually free to publish the results regardless of the study outcome. They
have no other conflicts of interest to declare.
Figure 1 Study design
Figure 2 Screenshot of the information screen used in the SC-IAT-P.
18
19